| GSM-R(Global System for Mobile communication-Railway)is a kind of mobile communication,which is specially applied to the operation and management of railway,and has become an indispensable part of carrying train control information transmission.In order to ensure the safe running of the train,operation and maintenance personnel need to know the running status of the train at all times.GSM-R interface monitoring system can analyze the interface communication data between devices,monitor and record the operation of GSM-R network in real time,and solve emergencies and failures problemWith the rapid development of railway technology in China,the amount of data transmitted by GSM-R network is increasing,and the storage and calculation of railway data by railway operators are also facing new challenges:the demand for analysis is constantly increasing,the data scale is constantly expanding,and the server of traditional centralized architecture is overwhelmed.It is imperative to establish a big data analysis platform to analyze and process mass data and quickly query statistics.At the same time,on this basis,GSM-R interference analysis and high-efficient location of the source has become another challenge to establish a stable and safe railway systemIn order to solve the above problems,according to the practical situation of the interface data of the GSM-R network system running in the current network,this thesis develops a big data platform of GSM-R interface monitoring based on Hadoop.Then,the interference data of air interface(Um interface)is analyzed.The main contents of this thesis are as followsFirst of all,this thesis deeply investigates and analyze the practical situation of Guangzhou Shenzhen Hong Kong line,and confirms that the massive data generated every day is totally beyond the capacity of the traditional centralized service architecture,so that the traction system adopts the distributed cluster architecture oriented to big data technology.Based on this demand,this thesis through technology selection,architecture analysis,module design to build a big data platform of GSM-R interface monitoring based on the Hadoop.Finally,through system test,verifies that the distributed cluster architecture can solve the shortcomings of the centralized service architecture in dealing with massive data,which has certain practical significanceThen,among all the monitored GSM-R interfaces,Um interface is the wireless interface,which is most vulnerable to all kinds of interference.The interference problem will seriously affect the quality of communication service of GSM-R network.By analyzing the interference sources,this thesis introduces several existing interference monitoring methods,which however have some limitations.Therefore,this thesis proves an interference analysis method which based on station spectrum modeling.By building the base station spectrum model and combining the spectrum data analysis,the B/S interference monitoring platform is realized,and the interference analysis results are displayed in real time through the graphical interfaceFinally,the platform has been successfully applied to Guangzhou Shenzhen Hong Kong line and Guangzhou Shenzhen line under the jurisdiction of Guangzhou Railway Bureau.The platform has formed a preliminary commercial scale,which not only effectively supports the needs of more than 30 front-end computers for real-time storage and analysis of big data platform,but also proves the effectiveness of the interface monitoring system in the current network operation through practical application cases. |
PDF Full Text Request